[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

RE: Solar Storms and Other Radiation Effects



Gunther,

Yes, there is evidence to suggest that ICs will achieve a slower rate of
threshold offset if their "bias" [a term the solid state physicists use]is
removed (i.e.,if they are turned off) as often as possible.  There is a
process called annealing which takes place as well, where some of the
trapped charge from charged particles bleeds off (find their way out of the
chip) and this will reduce the threshold offset.  But, the question remains,
how do you turn off the satellite fast enough to beat the onset of such an
event since it arrives so rapidly?

So, it is better if electronics not being used can be left off in orbit.
That may be much of the reason why AO-7 still lives.  It had a nice long
sleep up there.  During that time the IC may have literaly gotten better.
Rip Van Winkle and old wine come to mind.  :-)  A possible sensible use of
redundancy in a spacecraft would be to alternate between two redundant units
periodically where the spare unit is kept "cold" not "hot."

What I don't know is, for newer ICs, how much is the advantage in threshold
offset by keeping them off?

Jan

-----Original Message-----
From: Gunther Meisse [mailto:gogm@wmfd.com]
Sent: Friday, October 31, 2003 11:05 PM
To: Jan A. King
Cc: amsat-bb@amsat.org
Subject: RE: [amsat-bb] Solar Storms and Other Radiation Effects


Jan,
Thank you so much. I am sure other AMSATers enjoyed your information as much
as I have. It is great to hear such from those in AMSAT who have spent so
many years in the field...
Keep up the great job, professor.
Regards,
Gunther Meisse
W8GSM
P.S. I am interested in the matter of turning off the satellite prior to
such a storm. Since it is the accumulation of the radiation within the chip,
does turning it off make it any less subject to this accumulation?


-----Original Message-----
From: owner-AMSAT-BB@AMSAT.Org [mailto:owner-AMSAT-BB@AMSAT.Org]On
Behalf Of Jan A. King
Sent: Thursday, October 30, 2003 10:18 PM
To: amsat-bb@AMSAT.Org; rzee@utias-sfl.net
Subject: [amsat-bb] Solar Storms and Other Radiation Effects


Well,
I guess I was wrong.  There do seem to be people interested in this topic.
I got a lot of email about this from various sources.

First of all, have a look at these two pieces of data:

http://www.sel.noaa.gov/ftpdir/plots/2003_plots/proton/20031026_proton.gif

and

http://www.sel.noaa.gov/ftpdir/plots/2003_plots/proton/20031029_proton.gif

These are from the NOAA/Boulder website providing solar weather.  If you
like, you can look at this type of data over a many year period in this
level of detail.  It's all on this website.  What you have here is the
proton activity at geostationary orbit altitude over the period from October
24 thru October 30.  The value on the X axis is date, of course, and the Y
axis is the number of particles passing through every square cm of area (up
there) per second and from a solid angle in the sky of 1 steradian.  There
are 4*pi of those on the whole sky.  Each of the color plots is like passing
the particles through a high pass filter: the red curve is set at 10 MeV,
the blue curve is set at 50 MeV and the green curve is set at 100 MeV.  So
there are more particles with lower energy than higher energy and by orders
of magnitude.  Anyway, this is what's happened during the recent solar
event.  Look at the plot around mid-day on the 25th and you will see what
the background energy "noise" looks like.  Then the shock wave from the
solar wind hits the Earth late on the 26th of Oct. and pushes up the levels
levels orders of magnitude in a matter of minutes. Then it dies off over a
day or so and once again on Oct 28 at just before 1200 UTC the main event
hits the earth like a freight train.  It goes up two more orders of
magnitide in particle count and goes off scale. I chose the proton data
because I find it easier to interpret than other things on the website.
But, there is also electorn data, magnetic data, K-Index data and X-ray data
available on the same NOAA webite.  That data will show huge changes in
electron count and wild fluctuations in the magnetic field sign and
amplitude.  Also the so-called K-Index went to it's maximum value K=9 for
serveral days.  That's pretty unusual for sure.  So, if you are interested
in this kind of data this is the best website I'm aware of.  And, it has the
archive going back for many years.  To find out how unique this event was
(or was not) you can go back a few years to the solar maximum year and have
a look at data during the summer of 2001 and you will see quite a few events
similar to this one but, this was even larger.  I haven't tried to quantify
that part yet.  Anyway, this data gives you a feel for how fast things
change and how long they last up there.  Imagine plotting this on a linear
scale instead of semi-log!

Now note that this is what is happening at GEO orbit.  That's outside of the
radiation belts and in the beginning of the region known as
"flare-dominant."  AO-40 and the other Phase 3 spacecraft enter that region
at apogee (AO-40 for quite a pecentage of the orbit, actually).  But, LEO
orbits are partially shielded from these large dynamics and the belts will
introduce some time delay before the particle flux reaches these lower
orbits.  There may well be data published of a similar nature for LEO
satellites with suitable particle detectors.  But, I'm sure the flux will be
a few orders of magnitude lower than this GOES data.  There are regions in
the belts which trap protons and electrons and at various energy values.  To
get the entire picture you would have to visit all space around the earth in
lat/lon/altitude and at each point plot the energy spectrum and particle
count (like in the GOES data).  That would give you a complete instant snap
shot.  Then all of this changes with time.  So, the big picutre here is
complex.

Mostly, it's necessary to differentiate between dynamic processes like the
latest storm here and steady state processes which also affect our
satellites as well.

According to Paul Willmott VU9MU, AO-40 experience a large number of single
even upsets during the days surrounding this storm.  SEUs that are easily
detected and corrected occur in RAM memory and are "flips" in state caused
by a single particle colliding with a memory cell.  Usually protons with
energies on the order of 10 MeV have the highest probability of causing such
an event.  For elliptical orbits like Phase 3 satellites enhabit, the SEUs
occur near perigee where there are more higher energy protons during stable
field times.  But, apparently we are seening many of these events up near
apogee which can only mean that solar flares are causing them.  I'm not
discussing Single Event Latchups (SELs) here as that is a controversial
subject and even more complex.  Suffice it to say we haven't seen any of
those in our spacecraft so far.

But, what is also happening is that all IC's (and particularly MOS
transistor technology devices)accumulate charge damage from particles that
continually smack into them.  This cumulative dosage effect slowly biases
the switching threshold of the devices.  So, if a 5 v CMOS part normally
switches it's output state at 2.5 V going from say, a "0" to a "1" then, the
cumulative dosage it receives may cause it to switch at a higher values, say
3.0 V, then 3.5 V, until the device won't switch reliably anymore as the
switching threshold approaches the full 5.0 volt level.  When a flare occurs
it accelerates the cumulative dosage for awhile too but, it's the integral
over time that counts for that failure mechanism and that is largely
dependent on the orbit of the spacecraft.  Thus, for LEO satellites,
particularly those in lower orbits, this storm may not have prolonged
effects because of the shielding provided by the radiation belts (really the
magnetic field of the Earth).  Assuming our communications satellites don't
have fancy sensors then, if we survive the SEUs then the worst result is
that we have somewhat accelerated the accumulation of total dose.  [All of
this makes it so surprising that AO-7 still lives as it has spent about 30
years now in a pretty bad orbit (1450 km, circular, polar) and it is full of
first generation CMOS ICs.  Hundreds of them and most of them seem to still
be working.  AMAZING!].  If, like the Canadian MOST spacecraft there is a
sensor involved then it needs to be known what the sensitivity of the active
portion of the sensor (CCD or other) to various particles and at various
energies.  I don't know how to be definitive about such sensor damage
without having pretty good test data about the damage mechanism.  What can
be said is this last solar flare was pretty close to a 3-sigma event, which
says not too many more of these things will take place during this solar
cycle.  The best I could recommend is that the NOAA GOES data is a good
stable reference source against which one can take action.  But, because the
storm comes on so fast (as you can easily see here) there may be very little
time to act if you want to shut things off.  Look at the onset of the event
on October 28 and ask yourself if you were controlling a satellite would you
have had sufficient time to command your LEO satellite sensor system off
before the thing hit your spacecraft?  Usually small satellite systems must
wait until the spacecraft is in view of a ground station before action can
be taken.  Space systems like SOHO which have similar sensors and are
located at the L1 L'grange point will provide about 1 hour of waring before
the solar storm hits the earth and also there is some time provided if you
are in LEO because the mag field itself introduces some delay but, still
chances are you won't get to it in time.

In another email I can deal with the things we have done in AMSAT to shield
low cost spacecraft and what we have found are real issues and what we have
ignored and gotten away with.  But, I don't have time to do that right now.
I'll be going away until Monday night so, next week sometime, if there is
interest.

Jan VK4GEY/W3GEY
----
Sent via amsat-bb@amsat.org. Opinions expressed are those of the author.
Not an AMSAT member? Join now to support the amateur satellite program!
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org
----
Sent via amsat-bb@amsat.org. Opinions expressed are those of the author.
Not an AMSAT member? Join now to support the amateur satellite program!
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org



AMSAT Top AMSAT Home